[ https://issues.apache.org/jira/browse/KAFKA-2477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14735015#comment-14735015 ]
Jiangjie Qin commented on KAFKA-2477: ------------------------------------- [~hakon] Yes, that's correct. The log appending does the following two things: 1. Append message to log 2. Update Log.nextOffsetMetadata.messageOffset. If two follower reads come between 1 and 2. There will be a out of range exception. I think the fix is to read up to Log.nextOffsetMetadata.messageOffset for replicas instead of max size. Are you interested in submitting a patch? > Replicas spuriously deleting all segments in partition > ------------------------------------------------------ > > Key: KAFKA-2477 > URL: https://issues.apache.org/jira/browse/KAFKA-2477 > Project: Kafka > Issue Type: Bug > Affects Versions: 0.8.2.1 > Reporter: HÃ¥kon Hitland > Attachments: kafka_log.txt, kafka_log_trace.txt > > > We're seeing some strange behaviour in brokers: a replica will sometimes > schedule all segments in a partition for deletion, and then immediately start > replicating them back, triggering our check for under-replicating topics. > This happens on average a couple of times a week, for different brokers and > topics. > We have per-topic retention.ms and retention.bytes configuration, the topics > where we've seen this happen are hitting the size limit. -- This message was sent by Atlassian JIRA (v6.3.4#6332)